| Real-time database is a combination of real-time system and database technology,has the characteristic of high real-time, high data throughput and so on, and has beenwidely used in the field of industry automation control. Historical database is part of thereal time database, provides the functionality of storage and query historical data, andcan save the historical data for longer time. However, due to the large number ofcollection points, and the short period of acquisition time, it has to save a very largeamount of historical data. If these data are stored directly, the system will not only takeup a lot of storage space, but also will reduce the response time of the database andmake data transmission and query difficult.Therefore, data compression technologies are used in the historical database, it canmaintain the high real-time features of database, and at the same time improve thecapacity of the database as much as possible. Currently, the data compressiontechniques used in the historical database has mainly two kinds, lossless compressionand lossy compression. Lossless compression is generally based on the universalcompression theory, using the classical algorithms, such as Huffman compressionalgorithm; and lossy compression considers more to the characteristics of industrialreal-time data, taking some special care points algorithm.Research background of this paper is the research and development ofpetrochemical version of DCS configuration software. The main work is to study thevarious compression algorithms in-depth from the two sides of the data losslesscompression and lossy compression technology, and compare the advantages anddisadvantages of their performance, then design an efficient data compression schemeused in petrochemical version of DCS configuration software data compression on thehistory database, based on depth analysis of the characteristics of various types ofhistorical data. After comparing the advantages and disadvantages of various lossycompression methods, this paper also designs a new type of lossy data compressionalgorithm named ST(Step Trending) based on the linear trend of thought. And the newalgorithm is applied to the compression scheme, and achieves high performance. |
